HIV infection of the central nervous system (CNS) can cause severe problems including AIDS-related dementia. However, the brain excludes many useful anti-HIV drugs as a consequence of the blood-brain-barrier (BBB). This project attempts to improve the brain penetration properties of compounds like F-ddA, the prototype drug in this series, and a compound which began an NCI-sponsored anti-HIV clinical trial in June. The approach being explored is the design and synthesis of prodrugs with improved CNS transport properties which are enzymatically converted to an active anti-HIV agent after crossing the BBB. The 15 prodrugs synthesized during this project are 6-substituted-2'-fluorodideoxypurine nucleosides. These lipophilic compounds can be converted by adenosine deaminase (ADA) to 2'-beta-fluorodideoxyinosine (F-ddI), a compound with known anti-HIV activity. The fluorine atom, the distinguishing difference between F-ddI and the clinically useful drug, ddI, adds several practical benefits such as acid-stability for ease of oral administration, and reduced enzymatic catabolism. Our initial preclinical studies indicate that the 6-chloro analog may possess the best combination of the properties required to produce an effective CNS anti-HIV agent. This compound has been administered to rats in a comparison with ddI. Drug CSF and plasma levels, as well as brain tissue concentrations, are being determined.